Hi, I found a peculiar behavior in atoi() which isn't explained in atoi(3) - at least not in release 3.35 of the manual I have in Fedora 17. atoi() returns an int, but I expected that I can do unsigned int i = atoi(...); and it will just work fine, for any number - including numbers between 2^31 and 2^32. Any naive implementation of atoi(), repeatedly multiplying by 10 and adding the next digit, would have worked in this case just fine. But unfortunately, glibc's atoi() actually does something different in this case - for numbers >=2^31, it returns 2^31-1! The line above actually returned the wrong result on half the unsigned integer range :( I tested this on a 32-bit installation on Linux, by the way - I don't know if this makes a difference. I think it should be clearly state in atoi(3) what it does for numbers larger than 2^31 (or however this limit is actually defined in the code), and perhaps even explain why. I think it it should also be noted that atoi() is therefore unsuitable for reading unsigned ints - and for this purpose strtoul() should be preferred. Finally, atoi(3) states that "[its] The behavior is the same as strtol(nptr, (char **) NULL, 10); except that atoi() does not detect errors." In my test, this is NOT true - for numbers between 2^31 and 2^32 return different things in both functions (the correct number in strtoul, or 2^31-1 for atoi), and neither returns an error. -- Nadav Har'El | Monday, Jul 2 2012, 13 Tammuz 5772 nyh@xxxxxxxxxxxxxxxxxxx |----------------------------------------- Phone +972-523-790466, ICQ 13349191 |Why aren't fishmongers generous? Their http://nadav.harel.org.il |business makes them selfish. -- To unsubscribe from this list: send the line "unsubscribe linux-man" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
